In conjunction with the San Diego State University Global Change Research Group and staff of the San Diego State University Field Station Programs, the HPWREN team demonstrated the feasibility of high performance data communication between research air plane and a ground station during an approximately 15 minute flight. The communication substrate utilized unlicensed digital spread spectrum data radios, interoperable with Internet technology.
The objective was to assess an opportunity of supporting researchers and others on the ground in real time with inexpensive and freely available data communications means. An initial application was considered to be aerial support with high-resolution photography in real-time. There are varieties of other opportunities, including telemetry information from sensors on the plane, or ground information from sensors or other data sources, as relayed via an airplane.
This initial experiment consisted entirely of assessing the feasibility of integrating such wireless connectivity into fixed-wing airborne vehicles, and measuring the communication signal performance of the radio-to-radio connection. No user data exchanges were part of this test.
The location selected for this initial run was Ramona. A researcher home location closeby, with view of the Ramona valley and the Ramona airport and good networking connectivity, allowed for an easy and undisturbed installation of the ground station. Active participants of the demonstration included the pilot of the airplane, Rommel C. Zulueta and the principal technology integrator Pablo Bryant, both of the SDSU Field Station Programs. Additional HPWREN participants from UCSD included Todd Hansen and Hans-Werner Braun. A number of non-attending collaborators had expressed interest beforehand in the outcome of our experiments.
The distance between the ground station and the starting point of the airplane at the airport was approximately 3 miles. The radio on the airplane used a 30mW PCMCIA card, utilizing a Lucent COR. A 14dBi yagi antenna was pointing downwards from the belly of the airplane, while interfacing to the radio card via a bidirectional amplifier with 1 Watt output, and 17dB gain for the receiver. The total length of the yagi antenna is almost 19 inches, about 6 inch of which were extending below the frame of the airplane.
A full set of documenting photos of the event can be found at http://hpwren.ucsd.edu/Photos/20020409/.
Airplane and covered area
The airplane used is a single engine Sky Arrow 650 TCN ERA (Environmental Research Aircraft), utilizing a rear-mounted pusher propeller, part of the Global Change Research Group at San Diego State University.
The area covered extended between the Ramona Airport and Mt. Woodson.
About ten days days prior to the flight, Pablo Bryant and Hans-Werner Braun measured the performance of the metal frame surrounding the yagi antenna. Photos are available to document the activity. The signal and noise levels, as reported by the local and remote radios were measured as a function of the antenna length outside of the metal frame, while a connection was in place with a radio on top of Mt. Woodson:
local remote dstnce S N S N ====== ========== ========== 0 84 22 87 56 1 78 16 81 54 2 80 21 82 55 3 85 17 87 57 4 84 26 84 55 5 82 27 84 61 6 85 29 87 56 na 97 36 97 56 ====== ========== ==========
This showed a partially visible antenna will still create a viable connection. Since the pilot was only allowing us to go up to six inches outside of the air frame, special note was taken of the signal performance at that value.
Installation at the day of the event
During the actual event, the buffering battery, 5 and 12 volts power regulators, radio, power injector, and the antenna were mounted to the metal frame and bolted to the bottom of the airplane, after removing the passenger seat to gain the needed space.
In addition, a laptop was installed on top of the antenna frame and connected to a GPS receiver to log time stamped GPS data, that was later correlated to the radio signal performance as reported on the ground.
The actual flight
The ground station was set up with two polarized antennas connected to different radio cards, and mounted onto a tripod to manually tack the air plane during the flight.
Result data analysis
The GPS information on the airplane, and the radio signal performance measurements on the ground, were written into files for subsequent correlation and analysis. The GPS data was time stamped by the GPS receiver, while the ground station relied on Network Time Protocol data to synchronize its clock. This was by far precise enough, given that the GPS time resolution of its output data is not including any sub-second information.
The following graphs illustrates the times where there was a reported connection between the ground station and the airplane, as a function of time and altitude of the aircraft. The different colors represent areas of continuous connectivity, a color change means there was a known disconnect.
At the same time scale, the signal and noise property, as reported by the ground radio, as well as the signal to noise ratio are illustrated.
A further graph illustrates signal minus noise, comparable to what the radio management tool displays as SNR based on the signal and noise data.
The flight originated and ended at the Ramona airport, and geographical pattern were drawn to illustrate the areas of connectivity. In the following larger graphics, the researcher house, from which the measurements were made, is in about the vertical middle and horizontal left side of the images. The next image which focuses on that area, shows the ground station location approximately in the middle (northeast of the house).
The full flight, as reported by the GPS receiver on board of the airplane includes the airport location itself.
The next images show flight coordinates for the times where the ground radio had actual connectivity with the airplane, as a result of collected SNMP information.
Times where SNMP data was collected, with the radio link being down, are illustrated in the following:
The data points in the previous two images do not add up to the images displaying all GPS data point, due to periods of time where no SNMP data became available, not allowing for an indication as to whether the ground-plane link was online or not.
The result data shows that ground-air-ground communication using 802.11b spread spectrum radios is viable as an option to support researchers and incidents on the ground. Though tracking the air plane across 360 degrees with an antenna on a tripod was a challenge.
Next steps need to include gaining a better understanding for what antenna beam patterns we really need to support a signal cone below an airplane, if possible without having to steer the antenna. It would also be useful to experiment with different antennas, such as small patch antennas that would be more easily mounted to the frame of an airplane. A smaller setup would be highly desirable, taking up little space on the aircraft. Examples include small antennas and a small radio, amplifier, and power source.